Structural Basis of the Green–Blue Color ... - ACS Publications

Alexey Potapov , Wai-Ming Yau , Rodolfo Ghirlando , Kent R. Thurber , and Robert Tycko. Journal of the American Chemical Society 2015 137 (25), 8294-8...
3 downloads 0 Views 2MB Size
Subscriber access provided by UNIV OF HOUSTON

Article

The Structural Basis of the Green-Blue Color Switching in Proteorhodopsin determined by NMR Spectroscopy Jiafei Mao, Nhu-Nguyen Do, Frank Scholz, Lenica Reggie, Michaela Mehler, Andrea Lakatos, Yean-Sin Ong, Sandra J. Ullrich, Lynda J. Brown, Richard C. D. Brown, Johanna Becker-Baldus, Josef Wachtveitl, and Clemens Glaubitz J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/ja5097946 • Publication Date (Web): 21 Nov 2014 Downloaded from http://pubs.acs.org on November 26, 2014

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of the American Chemical Society is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

The Structural Basis of the Green-Blue Color Switching in Proteorhodopsin determined by NMR Spectroscopy Jiafei Mao1, Nhu-Nguyen Do1, Frank Scholz2, Lenica Reggie1, Michaela Mehler1, Andrea Lakatos1, Yean-Sin Ong1, Sandra J. Ullrich1, Lynda J. Brown3, Richard C. D. Brown3, Johanna Becker-Baldus1, Josef Wachtveitl2 and Clemens Glaubitz1*

(1) Institute for Biophysical Chemistry & Centre for Biomolecular Magnetic Resonance, Goethe-University Frankfurt, Germany (2) Institute for Physical and Theoretical Chemistry (3) Department of Chemistry, University of Southampton (*) Corresponding author. Email: [email protected] Institute of Biophysical Chemistry Goethe University Frankfurt Max-von-Laue-Str. 9 60438 Frankfurt am Main Germany Tel.: +49-69-798-29927 Fax.: +49-69-798-29929

ACS Paragon Plus Environment

1

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 29

Abstract Proteorhodopsins (PRs) found in marine microbes are the most abundant retinal-based photoreceptors on this planet. PR variants show a high level of environmental adaptation as their color is tuned to the optimal wavelength of available light. The two major green and blue subfamilies can be interconverted through a L/Q point mutation at position 105. Here, we reveal the structural basis behind this intriguing color tuning effect. High-field solid-state NMR has been used to visualize structural changes within green PR directly within the lipid bilayer upon introducing the green-blue L105Q mutation. The observed effects are localized within the binding pocket and close to retinal carbons C14 and C15. Subsequently, MAS-NMR sensitivity-enhanced by dynamic nuclear polarization has been applied to determine precisely the retinal structure around C14-C15.

Upon mutation, a

significantly stretched C14-C15 bond, a deshielding of C15 and a slight alteration of the retinal chain’s out-of-plane twist has been observed. The L105Q blue switch therefore acts locally on the retinal itself and induces a conjugation defect between isomerization region and imine linkage. Consequently, the S0-S1 energy gap increases resulting in the observed blue shift. The distortion of the chromophore structure also offers an explanation for the elongated primary reaction detected by pump-probe spectroscopy, while chemical shift perturbations within the protein can be linked to the elongation of late photocycle intermediates studied by flash photolysis. Besides of resolving a longstanding problem, this study also demonstrates that the combination of data obtained from high field and DNP-enhanced MAS-NMR together with time-resolved optical spectroscopy enables powerful synergies for in-depth functional studies of membrane proteins.

Keywords:

proteorhodopsin, retinal, color tuning, solid-state NMR, dynamic nuclear polarization, optical spectroscopy, pump/probe spectroscopy, flash photolysis

ACS Paragon Plus Environment

2

Page 3 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Introduction The surprising discovery of proteorhodopsin (PR) represented the first evidence of a bacterial retinal-based photoreceptor.1 It shows the typical structural scaffold of 7 transmembrane helices with the cofactor linked to the protein via a Schiff base and forms large pentameric and hexameric complexes.2,3 Its identification through meta-genomic screens of uncultured sea samples lead to many hundreds of PR-like sequences distributed in numerous microorganisms from different geographic areas.4-6 Their prevalent occurrence in microbial communities in the ocean’s photic zone and their ability to act as light-driven proton pumps7, which creates a transmembrane electrochemical gradient, makes retinal-based phototrophy a very important bioenergetic factor in marine ecosystems during nutrient deficient periods.8,9 PR has been extensively studied through concerted application of advanced biophysical methods including e.g. time-resolved optical spectroscopy10, AFM3, Raman- and infrared

spectroscopy11,

liquid12-

and

solid-state

NMR13,14,

DNP15,16,

EPR16

and

X-ray

crystallography17 as recently reviewed by Bamann et al.18 One of the most intriguing properties of the proteorhodopsin family is their high level of environmental adaptation with respect to optimized absorption of the available light. Two main PR subfamilies have been discovered in nature, which differ significantly in their light absorption profiles. The green light absorbing PR (GPR, max=525 nm) is mainly distributed in microbes living at the water surface, whereas PR absorbing blue light (BPR, max=490 nm) dominates at greater depths.19 The high sequence similarity between the differently colored PR variants makes this protein family an excellent case for studying general principles of color tuning, as the number of influencing residues are minimized. Understanding the factors controlling spectral tuning is a long-standing problem, which has triggered significant research efforts and has attracted additional attention with the emergence of optogenetics and the desire for engineering retinal-proteins with defined optical properties.20 Previous genomic analysis and biochemical studies have demonstrated that a single residue at position 105 serves as a major determinant for PRs wavelength regulation, which is a leucine in green and a glutamine in blue PR (Fig. 1). GPR can be switched into BPR with a single L105Q point mutation and vice versa. 19,21 This color-switching is also associated with a ten-fold slower photocycle in BPR, which has been suggested to correlate well with the reduced photon flux rate at the depth at which the blue PR gene was found. 22 A number of attempts were undertaken to explain the mechanism behind this distinct color tuning effect.23-28 Solution-state NMR and X-ray crystallography have shown that in both GPR and BPR the side chain of residue 105 is in close proximity to methyl group C20 at the end of the polyene chain.12,17 The same observation was made for bacteriorhodopsin (BR) in which the corresponding residue L93 shows a similar location and has also been shown to have an effect on color. 29 Replacing the non-polar side chain of Leu with the more polar one of Gln increases the polarity within the retinal binding pocket but decreases the molar partial side chain volume only slightly.30,31 Interestingly, a

ACS Paragon Plus Environment

3

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 29

recent extensive mutation study demonstrated that max shows no correlation with the polarity but with the side chain volume of amino acids at position 105, which does not explain the blue-shift induced by L105Q.28 Therefore, the mechanism behind the naturally occurring green-blue color-switch mutation remains unclear. To resolve this problem, methods have to be applied, which allow site-resolved monitoring of mutation effects and enable the detection of very fine structural perturbations within the chromophore-protein complex at sufficient resolution. Such data, which can be provided by solid-state NMR, will also help bridging the gap between quantum chemical and experimental approaches towards color tuning. Solid-state NMR offers a versatile approach to analyze structure, dynamics and functional mechanisms of membrane proteins embedded within lipid bilayers.32 In particular, the retinal protein field has benefited from cutting edge solid-state NMR developments mainly based on magic angle sample spinning (MAS-NMR). It has been shown that the finest perturbation in the sub-Å range within the retinal cofactor can be determined to high precision as demonstrated for bovine rhodopsin and bacteriorhodopsin.33-36 The availability of high field strengths offered the possibility of extensive resonance assignments on extensively labeled samples needed for further structure and dynamics analysis.14,37,38 In favorable cases, even a 3D structure determination is possible as shown for Anabaena sensory rhodopsin.39 The combination of dynamic nuclear polarization (DNP) with MASNMR has brought an essential improvement in sensitivity by orders of magnitude.40 This approach has proved to be especially useful for hypothesis-driven site-specific problems, which require work under cryogenic conditions as demonstrated for BR but also for other membrane proteins.15,41-43 In case of proteorhodopsin, solid-state NMR has made essential contributions including basic studies on retinal and protonated Schiff base44, unveiling the coupling between His75 and the primary proton acceptor Asp9713, secondary structure and dynamics analysis14,45,46 and resolving interactions between a distant loop mutation at position A178 and the retinal binding pocket15. Some of these studies are also powerful examples for the synergistic interplay between solid-state NMR (ssNMR) and optical spectroscopy and the high complementarity with X-ray crystallography.13-15 Here, we present an extensive solid-state NMR study on green PR (GPR) and its blue mutant GPRL105Q, which corresponds to the blue PR (BPR) occurring in nature. Multidimensional MAS-NMR experiments (PDSD, NCO, NCA, N(CA)CX) at very high fields applied to 13C, 15N-uniformly labeled samples allowed visualizing mutation-induced rearrangements within GPR via determination of chemical shift perturbations. Additional deuteration of GPR enabled the investigation of retinalprotein contacts via dipolar 1H-13C HETCOR experiments. DNP-enhanced MAS-NMR has been used to probe the effect of the green-blue switch on the retinal structure itself. The C14-C15 bond length and the HCCH torsion angle at the C14-C15 bond were determined using double-quantum spectroscopy. All of our data converge onto a molecular picture in which the interaction between Q105 and the retinal causes a distorted chromophore structure with effects on the conjugated -system, which alter the S0/S1 states resulting in a blue shift. Our findings will be discussed in the light of

ACS Paragon Plus Environment

4

Page 5 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

additional time-resolved optical data obtained in the time window from fs – ms showing a slower primary reaction and elongated lifetime of late photocycle intermediates.

Results Stationary and time-resolved optical spectroscopy of GPR and GPRL105Q A series of pH dependent absorption spectra of GPR and GPRL105Q solubilized in DDM have been recorded (Figs. 1b, c). At pH 9, a mutation-induced 20 nm blue shift from max = 520 nm to 500 nm is observed. The absorption maximum max depends on the protonation state of the primary proton acceptor D97.47 For GPR, max shifts from 520 to 545 nm (=25nm) when lowering the pH from 10 to 5. GPRL105Q follows this trend but max changes over a wider range from 537 to 500 nm (= a

of the primary proton

acceptor D97, which can be obtained from the inflection point of the sigmoidal titration curves and was found at 7.05 for GPR and 7.45 for GPRL105Q. We have probed the dynamics of the primary photoreaction by time-resolved optical pumpprobe spectroscopy in the 100 fs - ns time-range. The photo-isomerization around the C13=C14 double bond and the formation of the K-intermediate occur on this timescale. For comparison, time traces for three representative wavelengths where the signal is either dominated by the excited state absorption (ESA, 464 nm), by the generation of the photoproduct (PA, 556 nm) or by the stimulated emission (SE, 820 nm) are shown for GPR and GPRL105 in Fig. 2a. Overall, a similar behavior is observed for both samples, but the L105Q mutation slows down the primary reaction in a complex way. The data could be simulated with a sum of four exponential decays (Tab. 1). Flash-photolysis was then used to analyze mutation-induced alterations in the slower photocycle dynamics on a timescale of 1 µs to 20 s after photoexcitation. Transient absorption changes for GPR and GPRL105Q after photoexcitation are shown in Fig. 2b. At 590 nm, a decrease of the positive absorption signal in GPR is observed due to the decay of the K-intermediate, which is the photoproduct of the primary reaction observed in the pump-probe measurements mentioned above. This decay is accompanied by an increasing absorption monitored at 400 nm caused by the formation of a deprotonated Schiff base species (M-intermediate). The subsequent formation of the late N/O intermediates is indicated by an absorption increase at 590 nm. These species decay simultaneously, while the signal at 510 nm grows, which indicates the repopulation of the initial state. In contrast, the photodynamics of GPRL105Q is strongly altered. The amplitude at 560 nm is almost constant up to 100 s, which means that the K-state has decayed before and is not detected within the time window accessible by flash photolysis. Interestingly, the M-state observed through absorbance changes at 370 nm is not populated earlier, but its lifetime is significantly reduced. The subsequent formation of the late N/O intermediates at 560 nm is also strongly affected showing a complex behavior with an extended lifetime. This is further confirmed by the dramatically delayed repopulation of the initial

ACS Paragon Plus Environment

5

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 29

ground state monitored at 480 nm. A global fit analysis revealed that the data could be described by five exponential functions. A comparison with GPR is given in Table 1. Mutation-induced chemical shift changes in GPR by 13C-13C and 13C-15N MAS-NMR In order to identify structural changes induced by the green/blue L105Q mutation, we have characterized GPR and GPRL105Q embedded in lipid bilayers by MAS-NMR at high field (01H = 850 MHz). For chemical shift assignment, several dipolar-based 13C-13C and 15N-13C spectra on uniformly13

C,15N labeled samples were acquired. The aromatic residues are abundant in integral membrane

proteins and often show highly overlapping side chain signals. Therefore, in order to simplify spectral analysis, signals from Phe, Tyr, Trp as well as from two other abundant and hydrophobic residues (Leu and Val) were suppressed by reverse labeling. As shown in Fig. S1, the majority of peaks in 13C13

C PDSD and NCA spectra of GPR and GPRL105Q are superimposable. This indicates that these two

proteins have highly similar secondary structure. Using the known chemical shift assignment of GPR 13

GPRL105Q were unambiguously identified in

45

as a starting point, a number of residues in

13

C- C PDSD, NCA and N(CA)CX spectra, which

enabled the analysis of mutation-induced chemical shift differences (Tab. S1). Most of the observed changes are smaller than 0.2 ppm, which means that no major mutation-induced alterations in secondary structure or side chain conformations occur. However, a small set of 10 residues (T69, E85, T101, I112, A115, A116, I145, A185, T188, I194) has been found to show unambiguous

13

C chemical shift changes on side chain and/or backbone larger than the average

chemical shift differences of 0.4 ppm. Available structural data of different PR variants based on X-ray crystallography or liquid-state NMR reveal that residues T69, T101 and I194 are located within or close to the retinal binding site

12,17

. Example spectra showing significant chemical shift

changes for T101 and T69 are presented in Fig. 3a and b. The largest chemical shift change (0.8 ppm) was observed for C of the T101 side chain, which comes close to the retinal. In contrast, only small chemical shift changes have been found for the side-chains of D97 and D227, which serve as the primary proton acceptor and counter ions to the protonated Schiff base (pSB) (Fig. 3d). The 15N chemical shift of the protonated Schiff base at pH 9 (181 ppm) is also not affected by the L105Q mutation (Fig. 5a) and a comparison with the corresponding values from Schiff base counter ion model complexes and their correlation with max shows that GPR deviates from the free retinal systems more than GPRL105Q (Fig. 5b). Furthermore, the 15N chemical shift of N2 of His75, which forms a unique cluster with D97,13 remains unchanged (162 ppm, Fig. 5a). These data show that the L105Q mutation mainly influences residues in close proximity to the retinylidene chromophore. The locations of all residues for which a significant mutation-induced chemical shift change has been observed are highlighted in the GPR topology plot fesin Fig. 4.

ACS Paragon Plus Environment

6

Page 7 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Mutation-induced alterations of chromophore-protein contacts in GPR by 1H-13C MAS-NMR The mapping of mutation-induced chemical shift changes described above excludes aromatic residues, which are however found in the retinal binding pocket of PR. In order to detect these residues as well as to probe protein-retinal contacts, we have used a method based on the concept of selective interface detection (SIDY) originally designed for observing protein-ligand contacts by solidstate NMR.48 Instead of combining a 13C labeled ligand and unlabeled protein as used in the original version of SIDY, we have adopted an alternative labeling scheme in which unlabeled retinal is bound to uniformly 13C, 2H-labelled proteoopsin. Performing a dipolar 1H-13C HETCOR experiment on such a sample would in principle return a number of 1H-13C cross peaks as evidence for retinal-protein contacts. This approach is straightforward, as it does not require isotope labeling of the ligand by chemical synthesis. The 13C, 2H-labelled samples were reconstituted into highly deuterated lipids and kept in deuterated buffer in order to minimize the proton background. A long contact time (10 ms) in 1

H-13C CP was used for selecting long-distance contacts, including those crossing the protein-ligand

interface (Fig S2a). Small contributions from remaining protons resulting in residual short-distance 1

H-13C contacts within proteoopsin were eliminated using the MELODI scheme (Fig. S2b)

resulting 1H-13C MELODI-HETCOR spectrum is shown in Fig. 3c. The

49

. The

13

C dimension arises from

proteoopsin and the 1H dimension from the bound retinal chromophore. The observed cross peaks in the 1H-methyl – 13C-aromatic region represent therefore long-range through-space contacts between protein and co-factor. The 13C chemical shifts at 112.4 and 114.3 ppm are characteristic indole carbons of tryptophan. Based on the X-ray crystallographic structures of proteorhodopsin (4JQ6) and bacteriorhodopsin (1C3W), we assign these peaks to the conserved contacts between carbons C2/3/3 of Trp197 and protons of the C19/20 methyl groups of retinal (Fig. S2c). These two peaks become significantly weaker in the spectrum of GPRL105Q whereas an intense peak at a 13C chemical shift of 127.9 ppm appears, which could be tentatively assigned to a C2 of a Trp ring. These observations indicate a slight rearrangement of the Trp ring position relative to the methyl groups on the polyene chain.

Mutation-induced structural changes within the chromophore detected by DNP-enhanced MASNMR Our data show that the green/blue L105Q mutation causes very specific effects within the retinal binding pocket close to retinal carbons C14 and C15. Therefore, two samples of U-15N GPR and GPRL105Q in which the retinal co-factor has been doubly

13

C-labelled at these positions were

prepared. Due to the low sample amount restricted by the availability of synthetically labeled retinal, signal enhancement is required for obtaining an acceptable signal-to-noise-ratio. Furthermore, NMR under cryogenic conditions is needed for the precise determination of dipole couplings. We have therefore used DNP in combination with MAS-NMR.

ACS Paragon Plus Environment

7

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 29

DNP requires sample doping with suitable polarizing agents and dispersing them in a glycerolwater matrix, which forms a glass phase under cryogenic conditions (100 K). The optimal conditions for achieving the best possible enhancement are sample dependent and require careful adjustment according to the nature of the samples. Here, the biradical TOTAPOL

50

was used and DNP

enhancement was monitored via 1H-15N-cross polarization. In an initial approach, TOTAPOL was directly added into the protein/detergent/lipid reconstitution mixture attempting to maintain an equal radical concentration inside and outside of proteoliposomes. This strategy was not successful since TOTAPOL was absorbed by biobeads used for detergent removal. We therefore soaked the proteoliposome pellet directly in the appropriate radical solution. Samples were incubated for some hours before transferring them into a MAS rotor. Without glycerol, best enhancement (18x) is achieved at 5mM TOTAPOL, while in the presence of 30% glycerol (d8-glycerol/D2O/H2O 30/60/10), an up to 30-fold enhancement at 20mM TOTAPOL was detected (Figs. 5c & d). Further increase of glycerol concentration did not provide any improved enhancement and was therefore avoided. A conventional 15N-CP MAS NMR spectrum of U-15N GPR is compared to a DNP-enhanced spectrum in Fig. 5a and 5c. As described above, a significant signal enhancement is achieved, but some line broadening due to the low operation temperature needed for DNP (100 K) is observed as well. The 15N signals of pSB and H75 are covered by a broad signal arising probably from the His-tag centered on 170 ppm. The His-tag is usually not observed in non-frozen GPR samples above pH 7.5 but crosspolarizes under cryogenic conditions. Since the sample carries 14-15-13C-all-trans retinal, the pSB signal can be cleanly filtered out by two cross polarization steps, in which first 1H magnetization is transferred to C14 and C15 and from there to the directly bonded

15

N of the pSB (Fig. 5e). The

remaining backbone signal arises from HCN double-CP transfers between 15N and natural abundance 13

C within proteorhodopsin. As shown in Fig. 5e, the chemical shift of pSB 15N in GPR is observed at

181 ppm, which is identical to room temperature measurements (Fig. 5a). This signal broadens but the identical chemical shift indicates an intact and functional state in line with earlier cryo-FTIR studies.51 DNP-enhanced 1H-13C CP MAS-NMR spectra of 14,15-13C-retinal bound to U-15N PR are shown in Fig. 6. A 27-fold enhancement using the conditions described above is achieved. Such an enhancement allows detection of the C14 and C15 signals even when using as little as 1 mg of protein sample and with a much reduced experimental time. All natural abundance signals are removed upon applying a dipolar double-quantum filter. The two remaining resonances are assigned to C14 and C15 of the retinal cofactor. For GPR, resonances of C14 and C15 are observed at 120 ppm and 161 ppm, respectively. The linewidths of these two peaks are 184 and 159 Hz, respectively, which are both larger than the upper limit of the homogeneous linewidth converted from refocused single-quantum coherence lifetimes (Fig. S3). In the L105Q mutant, chemical shift changes are observed as C14 shifts to 121.3 ppm and C15 to 165.4 ppm (Fig. 6b). In both constructs, the chemical shifts of these two sites (Tab. 2) clearly point to a single all-trans conformation of retinal in the ground state.

ACS Paragon Plus Environment

8

Page 9 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

We have further analyzed double-quantum built-up curves to determine the exact bond length between C14 and C15 (Figs. 7a-c). We have adapted a standard POST-C7-based DQ scheme and have monitored the signal intensity while varying the DQ excitation and reconversion steps simultaneously. 33

It has been shown that this method is able to precisely determine inter-nuclei distances and is rather

sensitive to small bond length alterations. For GPR, a dipole coupling of 2665 Hz corresponding to a bond length of 142 pm was found, which is significantly shorter than the 146 pm expected for an ideal single bond length in linear polyene systems as found in crystalline all-trans retinal.52 In contrast, a dipole coupling of 2450 Hz resulting in a bond length of 146 pm was determined for GPRL105Q. In addition to the C-C bond length, we have also determined the planarity around the C14-C15 bond by measuring the H-C14-C15-H dihedral angle. This is achieved by heteronuclear local field (HLF) experiments, which spectroscopically “isolate” this fragment from the environment and have been successfully applied on biological solids.53,54 Here we choose a recently released version of HLFHCCH experiments that uses twice the dipolar phase accumulation time of the original version.53 This method reduces systematic error by improving the angular resolution of the measurements and also raising the number of acquirable data points and is therefore better suited for determining the fine structural details required here. Experimental and simulated dephasing curves are shown in Figs. 7d and e. For GPR, a best fit is obtained for 161°, which changes to 164° in GPRL105Q. Therefore, this moiety adapts trans-like conformations in both proteins. However, the measured values still differ significantly from an ideal trans conformation by about 16 - 19o. This deviation, which does not exist in all-trans-retinal in solution, can be due to the steric effects from protein environment as also found in BR.54

Discussion Optical spectroscopy shows that the L105Q mutation causes a significant blue shift of 20 nm at pH 9 in green proteorhodopsin (Fig. 1b), as observed for native blue proteorhodopsin.19,21 This color tuning is accompanied by a slower primary reaction and extended lifetime of late photocycle intermediates (Figs. 2a & b). The approximately 10-fold slower photocycle of GPRL105Q resembles the characteristics of naturally occurring blue PR.

19,21,22

GPR therefore offers an excellent platform for

understanding the molecular basis of this mutation-induced color tuning. Detergent/lipid effects or an altered oligomeric state of GPRL105Q compared to GPR can be excluded as the source of the color shift as both samples were prepared and studied under identical conditions and both form oligomeric complexes of the same size (Fig. S4). Despite the known location of the mutation site in close proximity to methyl group C20 at the end of the polyene chain12,17 and numerous spectroscopic and biochemical studies23-28, the underlying mechanism of the observed color shift remained unclear, which emphasizes the need for using high-resolution and high-precision methods such as solid-state NMR to answer this key question.

ACS Paragon Plus Environment

9

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 29

Mutation-induced chemical shift perturbations visualize highly localized structural effects on GPR We have identified unambiguous mutation-induced chemical shift changes in a small number of residues. Their location within the topology plot is shown in Fig. 4. Most of them are found close to the mutation site in helices C and F. The small number of affected sites indicates a highly locationspecific effect. This is in contrast to another color-tuning mutation, A178R within the EF loop, which causes a large number of chemical shift perturbations starting from the EF-loop and spreading throughout the whole protein.15 The most pronounced chemical shift change is observed in the side chain of residue T101 close to the end of the retinal polyene chain (Fig. 8a). This residue is located just beneath L105 on the same side of helix C pointing towards the retinal chromophore (Fig. 8b). This observation reveals a significant coupling between both sites. Interestingly, this TxxxL motif (Fig. 1a) seems conserved, as it is not only found in PR but also in BR and XR.55 The chemical shift perturbations do not propagate further along helix C. The primary proton acceptor D97 located below T101 and the primary proton donor E108 above L105 are not affected by the L105Q mutation. It appears that the structural changes triggered by this mutation do not extend more than two turns towards the extracellular direction. Similarly, no significant perturbations are observed for the pSB K231 and for D227 (Fig. 8b). The occurrence of primarily short-range interactions is also seen for chemical shift changes of W197 close the L105Q site, which do not extend to M134 on the retinalfacing side of helix D (Fig. 8b). Interestingly, Q105 and W197 were shown to interact via H-bonded water molecules in some of the protomers of the blue PR X-ray structure.17

Structural basis of the green-blue switch in proteorhodopsin In general, color tuning, i.e. the opsin shift, depends on a number of mechanisms, which affect the delocalization of the positive charge at the protonated Schiff base along the polyene chain and the S0 - S1 gap. 56-58 Contributing factors include (i) the orientation of the cyclohexene ring, (ii) an altered counter ion – Schiff base distance and (iii) structural and electronic distortions of the retinal polyene chain itself caused by local steric or electrostatic effects. (i) Alterations in the cyclohexene ring/polyene chain coplanarity can contribute up to 20% to the opsin shift 59 and are mainly determined by altered protein – ring contacts. Such an effect does not seem likely in our case, since no chemical shift changes in residues surrounding the cyclohexane ring have been observed, which would be indicative of structural re-arrangements. This is in contrast to the red-shifting A178R mutation, which triggers chemical shift changes within residues surrounding the ring showing that the binding pocket is allosterically regulated by the distal EF loop.15 (ii) The electrostatic effect caused by an altered distance between the protonated Schiff base and the counter ion is a major color-determining factor. The 15N chemical shift of the pSB nitrogen is highly sensitive to the formation of H-bonds, the dielectric properties of its surrounding vicinity and the counter ion distances. It can be directly linked to max of the retinal pigment60,61 (Fig. 5b). However, the L105Q mutation does neither affect the chemical shift of the pSB nitrogen nor of its interacting

ACS Paragon Plus Environment

10

Page 11 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

residues D97 and D227, which are part of the counter ion complex. Moreover, H75, a residue involved in H-bonding with D97,13 remains unaffected and the pKa of D97 increases only slightly. Previous studies hypothesized that the blue-green switching in GPR is based on an altered pSB environment.19,26,28 Our data show that the net effect of the protein environment on the pSB nitrogen seems rather similar in GPR compared to its blue version. Therefore, altered counter ion effects are probably not a major determinant of the L105Q mutation induced blue shift. (iii) The mutation-induced chemical shift perturbations point towards a protein-induced structural rearrangement within the retinal close to the end of its polyene chain (Fig. 8a). To gain closer insight into alterations within this part of the chromophore,

13

C-labels were introduced at

positions C14 and C15 and samples were analyzed using DNP-enhanced MAS-NMR ensuring highest sensitivity, precision and localization. Upon mutation, a deshielding of carbon C15 (Fig. 6b) and a stretching of the C14-C15 bond length by 4 pm (Fig. 7c) was observed. Furthermore, we found a significant out-of-plane twist of the H-C14-C15-H torsion angle in both green and blue PR constructs (Fig. 7f). Previously, the localization of such a conjugation defect and its relation with the chemical shifts of the odd numbered carbons has been extensively studied and described for rhodopsin.

62,63

In

our case the observed C14-C15 bond stretching together with the deshielding of C15 show that the conjugation defect induced by the protonated Schiff base covers the end of the polyene chain. This defect could propagate further to the isomerization region as indeed indicated by the strongly altered isomerization kinetics upon mutation (Fig. 2a, Tab. 1). It is worth noting that previous studies in rhodopsin also support our finding that the conjugation defect could be solely regulated by the polyene chain rather than ionone ring and Schiff base. 64 The question is to which extent this defect is caused by steric or electrostatic interactions and how this would explain the observed blue shift of 0.1 eV. A glutamine at position 105 increases polarity and H-bonding options and reduces the side chain volume slightly. A previous theoretical study based on homology modeling and molecular orbital calculations suggested that Q105 could cause a blue shift by differently stabilizing S0 and S1-states if the side chain dipole moment points with the carbonyl oxygen towards the imine linkage.24 The authors also proposed that this orientation would be ensured by a hydrogen bond formed between the Q105 carbonyl group and the proton at the retinal carbon C15. Indeed, the deshielding of C15 observed here would be compatible with the formation of a hydrogen bond and the x-ray structure of blue PR also indicates the correct orientation of the Q105 side chain.17 The importance of the orientation of a polar side chain in the retinal binding pocket is illustrated by the fact that other computational studies have reported red shifts when placing a glutamine close to the retinal.65 The resulting bond stretching and its consequences can also be considered from the perspective of altered electronic configurations in the retinal pigment upon photoexcitation (Fig. 8c).6668

The highest occupied molecular orbital (HOMO) features a characteristic conjugated -orbital

pattern extending along the polyene chain with its maximum at the cyclohexene ring. This conjugated

ACS Paragon Plus Environment

11

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 29

-system shifts towards the pSB in the lowest unoccupied molecular orbital (LUMO). The energy gap between these two states defines the absorption wavelength. C14-C15 bears a single bond character in the HOMO but tends towards a double bond character in the LUMO. The observed elongation of C14C15 bond in GPRL105Q disrupts the delocalization of -electrons at this position in the LUMO and could therefore destabilize this state. In contrast, a stretching of the C14-C15 bond does not strongly affect the HOMO (Fig. 8c). A conjugation defect also hinders the dislocation of the positive charge causing an accumulation at the end of the chain, which is in line with the observed deshielding of C15. In the LUMO, the conjugated -system locates more closely to the pSB nitrogen, which further promotes the migration of pSB positive charge towards the electron-rich polyene part. A defect around the C14-C15 bond weakens this pathway more significantly in the excited state. As a consequence, the energy gap increases and a blue shift of the absorbed light is observed. The effect will be even stronger when considering that the conjugation defects spreads over a number of adjacent bonds into the isomerization region. The observed twisting of the polyene chain around C14-C15 is similar to bacteriorhodopsin.54 Since the L105Q mutation reduces the H-C14-C15-H torsion angle only slightly (Fig. 7f), its color tuning contribution is most likely negligible, also taking into account that a higher co-planarity would probably stabilize the excited state over the ground state resulting in a red and not a blue shift.59

Effect of the green-blue switch on the photodynamics of proteorhodopsin Aside from explaining the source of the observed blue shift, our NMR data also provide clues for understanding the mutation-induced alterations in PR photodynamics. The generally slower primary reaction in GPRL105Q (Fig. 2a) is most likely a consequence of a conjugation defect in that region, which results in a higher isomerization barrier. This is a direct consequence of the mutation-induced alterations of retinal structure and electronic configuration caused by surrounding residues such as Q105 and T101. The generation of photocycle intermediates is also strongly affected (Fig. 2b). The lifetime of the M-state is reduced indicating a faster reprotonation of the Schiff base during the photocycle. In contrast, the decay of the late N/O intermediates is significantly elongated, which accounts for the overall longer photocycle. The decay of these intermediates is associated with the reprotonation of the primary proton donor E108 in helix C, which is close to the mutation site, and involves an outward movement of helix F to support proton uptake from the cytoplasmic side as known from BR.69-72 Our data show that significant backbone chemical shift changes induced by this mutation extend towards the cytoplasm to A116 on helix C, which passes E108 (Fig. 4). Furthermore, it has been shown for BR that an outward movement of helix F hinged around a Trp residue.29,73 The corresponding Trp (W197) in PR is affected by residue 105 and this change spreads to other residues (A185, T188 and I194) on the cytoplasmic half of helix F. These local conformational rearrangements appear related to the altered lifetime of the N/O intermediate states in the blue PR mutant. Indeed, some mutation in these locally affected regions extend the

ACS Paragon Plus Environment

12

Page 13 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

lifetime of N/O state in PR significantly.74 A link between the O-state decay and proton pumping activity has been found by analyzing various proteorhodopsin variants 74: The slower the O-state decays the less efficient becomes proton pumping. This fact has been linked to the biological activity of BPR.22 Our data therefore provides a clear structural clue to this functional adaption.

Conclusions Due to their environmental light-adaptation, proteorhodopsins are an intriguing class of retinal proteins. Despite of their sequence diversity, their blue-green color switch is dominated by the nature selected single point mutation L105Q. This makes this group of proteins an ideal showcase for addressing the molecular mechanism of the phenomenon of retinal-based color tuning. By combining high-field MAS-NMR for high spectral resolution and DNP for unprecedented sensitivity with properly designed labeling schemes, we were able to probe the structural changes in both protein and retinal at site-specific level and high precision. Our data clearly show that the color-switch only triggers highly localized structural changes within PR, which can be correlated to the slowed photocycle and point to the end of the retinal molecule as the hotspot for the color tuning. Further computational approaches such as QM/MM simulations confined by MAS-NMR parameters will enable accessing the full electronic configuration of retinal and surrounding protein environment. Our study also presents an approach for obtaining valuable high resolution experimental structural information directly correlated to the optical properties of protein-chromophore systems, which could support and complement computational and biochemical/biophysical as well as other structural biology approaches on such important targets.

Materials and Methods Sample preparation GPR (eBAC31A08 variant1) expression and purification were carried out as described previously.46 The GPRL105Q gene was kindly provided by Prof. Spudich (The University of Texas Medical School). GPR and GPRL105Q were cloned in a pET27b-plasmid and expressed in E. coli C43 cells. U-[2H, 13C, 15N] GPRs were expressed in the same E. coli strain using a protocol modified from Ward et al.75. A five steps adaption (LB with 30, 60 and 90 % (v:v) D2O and M9 with 90 and about 99 % D2O) was performed before expression in perdeuterated medium. High kanamycin concentration (200 g/ml) was used in order to stably maintain the plasmid. The starting OD600 nm in perdeuterated M9 medium (3 - 4 g/L U-[2H, 13C] glucose and 1 - 2 g/L 15NH4Cl) was 0.14 and cells were incubated for about 10 h at 30 oC (220 rpm) until OD600 nm reaches between 0.6 and 0.8. IPTG was then added to reach a concentration of 200 mg/L for induction together with 35.5 L retinal solution (10 mg/ml in d6-EtOH). The culture was incubated for 12 h under the same condition and another batch of retinal solution was added (40 L). After that, the culture was continued for another 8 h before cell harvesting. The protein was solubilized in 1.5% n-dodecyl--D-maltoside (DDM) at 4°C overnight. The

ACS Paragon Plus Environment

13

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 29

solubilized protein (supernatant) was purified using Ni-NTA matrix. The bound protein was eluted with 500 mM imidazole in 0.05% DDM. Purity was checked by absorption spectroscopy and SDSPAGE. In addition, blue native PAGE analysis was carried out (Fig. S4). The protein yield for both constructs was about 11 mg/L. For solid-state NMR experiments U-[13C,15N] PR was reconstituted in DMPC:DMPA (9:1) liposomes in a protein to lipid ratio of 2:1 (w:w) as previously reported. The proteoliposome was collected by ultracentrifuge and washed intensively in NMR sample buffer containing 50 mM Tris and 5 mM MgCl2 at pH 9. U-[2H, 13C, 15N] PR was reconstituted into d67-DMPC liposome at the lipid-toprotein ratio about 20:1. The complete incorporation of PR into deuterated lipids was confirmed by sucrose gradient experiment. The deuterated sample was washed multiple times in fully deuterated NMR sample buffer and then incubated at 4°C for two weeks during which the buffer was changed at least 3 times. The pD of the deuterated NMR buffer was calibrated to be equivalent to pH 9.0 as used on protonated samples in this work. 14-15-13C-all-trans-retinal was synthesized as reported previously.15 It was incorporated into GPR by directly adding it to the medium as described before. Reconstituted samples were incubated overnight at 4°C with 20 mM TOTAPOL in a buffer containing 10 % H2O, 30 % d8-glycerol and 60 % D2O. The solution was removed carefully and the pellets were transferred to a 3.2 mm zirconium rotor. Each sample for screening DNP conditions contained 2.5 mg protein. For the measurements on 14-1513

C-all-trans-retinal labeled GPR, 2 mg of GPR and 5 mg of GPRL105Q were used.

Optical spectroscopy Stationary and time-resolved optical spectroscopy was carried out as described previously.

15

GPR and GPRL105Q were used solubilized in 0.15% DDM, 150 mM NaCl, 50 mM Tris, pH 9. The sample was diluted to an OD of 0.5 (d = 0.1 cm) for pump-probe spectroscopy and to an OD of 0.7 (d = 1 cm) for flash photolysis.

High-resolution MAS-NMR experiments All high-field MAS-NMR spectra for assignment were acquired using standard pulse programs on a Bruker Avance III spectrometer operated at 20 T (850 MHz as 1H Larmor frequency) equipped with a 4 mm triple-resonance probehead. The MAS frequency was stabilized at 10 kHz and the effective sample temperature was about 4 oC. The mixing time in PDSD experiments was 20 ms. The spectra acquired on GPR and GPRL105Q mutant were processed in the same way. The 1H-13C MELODI HETCOR experiment was performed using a pulse sequence similar to the one reported in literature.49,76 The MAS was set to 11.26 kHz. 1H-13C dipolar dephasing was achieved in two-rotor periods with 20 s dephasing time. The 1H FSLG homo-nuclear decoupling77 field strength during dephasing period and the chemical shift encoding period were 92 and 90 kHz respectively. Data acquisition in the indirect dimension was incremented by 72.6 s and in total 64 t1

ACS Paragon Plus Environment

14

Page 15 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

points were collected. A long 1H-13C CP with 10 ms contact time was used. SPINAL6478 decoupling with B1 field at about 50 kHz was applied during acquisition. The spectra of two constructs were scaled using a same factor (0.578).

DNP MAS-NMR experiments All DNP MAS-NMR experiments were performed on a Bruker Avance II wide-bore spectrometer operated at 392.78 MHz and equipped with a triple resonance 3.2 mm cryo-MAS probe head. A MAS frequency of 8000 ± 2 Hz was used for all experiments. The temperature inside the stator was kept at 100 K. Microwaves were generated from a gyrotron (Gycom, Russian Federation) operating at 259 GHz. Energy loss within the probe head are below 2 dB and the overall decay of microwave energy is about 5 dB from gyrotron to stator. The effective microwave power applied on our NMR samples was about 2 W. The 1H-13C/15N CP and 13C-15N DCP spectra were acquired using standard pulse sequences with SPINAL64 hetero-nuclear decoupling at 135 kHz during acquisition.78 The POST-C7 scheme was used for exciting and reconverting DQ (double-quantum) coherence.79 A continuous wave decoupling at 112 kHz and a SPINAL6478 decoupling at 105 kHz were applied during POST-C7 pulses and acquisition time, respectively. For recording DQ build-up curves, the durations of DQ excitation and reconversion periods were varied simultaneously by changing both the numbers of C7 excitation and reconversion cycles in steps of 142.9 s.33 Each step was recorded with 512 scans. The HLF-HCCH experiments were carried out as described by Levitt and co-workers 53: DQ coherences are excited, evolve under homonuclear proton decoupling and are detected after a reconversion step. Two complete POST-C7 cycles were used for both DQ excitation and reconversion. The 13C carrier frequency was placed in the middle of the resonating frequencies of C14 and C15. A PMLG (phase modulated Lee-Goldburg) homonuclear decoupling step 77 with an RF field of 112 kHz was applied during the DQ-evolution time, which was incremented by multiple integers of 18 PMLG cycles (1/8th of one rotor period). Two equal proton-decoupling periods (112 kHz) were applied before and after the evolution time in order to keep the total evolution time constant (two rotor periods, 250 s). Typically 1024 to 2048 scans were accumulated for each spectrum.

Spectral assignment, spin dynamics simulation and data fitting The analysis of the PDSD, NCA, and N(CA)CX spectra of GPRL105Q were carried out in CARA (cara.nmr.ch) using a reference chemical shift data set of GPR (BMRB 15955, 17817)14,45 as starting point. All assignments were validated in all spectra and are summarized in Tab. S1. For further analysis, only unambiguous assignments with chemical shift differences larger than 0.4 ppm were used. Spin dynamics simulations of DQ build-up and HLF-HCCH dephasing curves were performed using SIMPSON.80 The 1H-1H and 1H-13C decoupling was treated by turning off the 1H-1H and 1H-13C

ACS Paragon Plus Environment

15

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 29

dipolar interactions. The other details are included in the SIMPSON script in SI. A series of DQ buildup curves were calculated by varying the

13

C-13C dipolar coupling constant from 2150 Hz (1.52 Å

distance) to 3000 Hz (1.36 Å distance). To account for transversal relaxation, DQ build-up curves were multiplied with a mono-exponential decay function. Experimental data were analyzed by searching for global minima under variation of bond lengths and the time constant used to describe the exponential decay. Unambiguous minima were obtained for both experimental data sets (Fig. 5a). In order to simulate the HLF-HCCH data, a library of HCCH conformers was created by varying the dihedral angle is steps of 1° using PYMOL (Schrödinger LLC). The complete parameters used to describe the spin system are given in the supporting information. The appropriate C14-C15 bond lengths determined from the DQ build-up curves was used.

For each of the resulting HCCH

conformers, a dephasing curve was calculated. The experimentally observed transverse relaxation was accounted for by multiplication with a mono-exponentially decaying function. The best-fit solution was searched through a backward iterative optimization of both dephasing curve and exponential decay. Python and shell programming scripts were used to automate simulation and numerical optimization. As a control, additional simulations were carried out, which included an additional 15N and 1H at the Schiff base position. Their effects on the dephasing curves were negligible.

Supporting Information: Comparison of PDSD and NCA spectra of GPR and GPRL105Q (Fig. S1); Supporting data for 1H-13C HETCOR experiments on [U-13C, 2H]-GPR and GPRL105Q (Fig. S2); DNP spin-echo experiments on [14, 15-13C-all-trans-retinal]-GPR (Fig. S3); Blue-native gels of GPR and GPRL105Q in DDM micelles and liposomes (Fig. S4); Table of resonance assignments and list of chemical shift changes (Tab. S1); SIMPSON script or HLF-HCCH spin simulations. This material is available free of charge via the Internet at http://pubs.acs.org.

Acknowledgements: The work was funded by DFG/SFB 807 ‘Transport and communication across membranes’. The DNP experiments were enabled through an equipment grant provided by DFG (GL 307/4-1 and Cluster of Excellence Macromolecular Complexes Frankfurt). We are grateful to Igor Shapiro, MPI Chemical Energy Conversion, for helpful discussions and initial QM/MM calculations. Vasyl Denysenkov and Jörn Plackmeier (University of Frankfurt) are acknowledged for their help in establishing the DNP setup and in TOTAPOL synthesis. Marie Concistre and Malcolm H. Levitt are acknowledged for initial advises on the HCCH data analysis. Leonardo Gonnelli and Gianluca Gallo (CERM, Florence) are acknowledged for valuable and practical suggestions on protein perdeuteration.

ACS Paragon Plus Environment

16

Page 17 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Figures and Tables

Fig. 1: (a) A protein sequence alignment across the PR family shows a leucine for the green and a glutamine for the blue variant at position 105 as well as a conserved TxxxL/Q motif between residues 101 and 105. Residue 105 is located in the vicinity of the retinal cofactor 12,17. Important residues are highlighted in the cartoon (proton donor E108, proton acceptor D97, Schiff base K231, conserved H75 stabilizing D97). The alignment was generated using WebLOGO. (b) Stationary light absorption spectra of GPR and GPRL105Q at pH 9 reveal a mutation-induced blue shift of 20 nm. (c) A pH titration shows that the observed blue shift is largest above the pKa of the primary proton acceptor, which is 7.05 and 7.45 for GPR and GPRL105Q, respectively. The pH-dependent color change is more pronounced in GPRL105Q.

ACS Paragon Plus Environment

17

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 29

Fig. 2: (a) Time traces of the primary photoreaction for GPR and GPRL105Q at pH 9. The absorption changes at 464 nm, 556 nm and 820 nm are dominated by excited state absorption (ESA), photoproduct formation (PA) and stimulated emission (SE), respectively. Lifetimes are extracted from these curves by fitting four exponential decays (Tab. 1). (b) Laser-flash induced transient absorption changes at pH 9 for GPR and GPRL105Q on a timescale of 1 µs to 20 s. The transients are representative for the dynamics of the ground state population (510 nm / 480 nm), the K-decay at early delay times and the formation and decay of the N/O-intermediates at late delay times (590 nm / 560 nm) and the formation and decay of the M-intermediate (400 nm / 370 nm).

ACS Paragon Plus Environment

18

Page 19 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Fig. 3: The chemical shifts of residues close to site 105 and/or in the retinal binding pocket are significantly perturbed by the L105Q mutation. (a) The largest effect is observed for T101 as identified in 13C-13C PDSD (20 ms mixing time) and 15N-13C N(CA)CX MAS-NMR spectra while other Thr residues shown in the same spectral region are not affected. (b) Further effects are observed for example for T69 in a N(CA)CX spectrum and (c) for W197 in a 1H-13C HECTOR spectrum. (d) In contrast, D227 and D97, which are part of the counter ion complex to the protonated Schiff base, are not influenced (PDSD of 20 ms mixing time). Spectra are color-coded (green=GPR, blue=GPRL105Q). See text for further experimental details and labeling schemes used. ACS Paragon Plus Environment

19

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 29

Fig. 4: Topology plot with a summary of all observed chemical shift perturbation in GPR upon L105Q mutation. The mutation site is colored in blue, residues showing significant 13C chemical shift changes (> 0.4 ppm, T69, E85, T101, I112, A115, A116, I145, A185, T188, I194 and W197) are highlighted in yellow and some functionally relevant residues (proton acceptor D97, D227, proton donor E108, pSB residue K231) are colored in white.

ACS Paragon Plus Environment

20

Page 21 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Fig. 5: (a) 15N-CP MAS NMR spectrum of U-15N GPR recorded at 280 K. The characteristic 15N signals of the pSB (181 ppm) and N2 of His75 (162 ppm) are not affected upon L105Q mutation. (b) Correlation between max and pSB 15N chemical shifts obtained from retinal derivatives with all-trans polyene chains with different halide counter ions 81 in comparison to BR, GPR and GPRL105Q. (c) 15NCP MAS NMR spectrum of U-15N GPR recorded under DNP conditions at 100 K and 258 GHz / 393 MHz (e- and 1H Larmor frequencies). (d) Screening different sample conditions shows that a maximum of 30-fold signal enhancement is obtained using 20 mM TOTAPOL in 30/60/10 d8glycerol/D2O/H2O mixture. The DNP enhancement was determined from the 15N backbone resonance and calculated according to Imw on / Imw off. (e) DNP-enhanced 13C-15N double CP experiment on [14,1513 C-all-trans-retinal-U-15N]-GPR. In this experiment, magnetization is transferred from protons to 13Clabelled retinal carbons C14 and C15 and from C15 further to the 15N-labelled nitrogen of the pSB. Its resonance under DNP condition is slightly broadened but occurs at the same chemical shift compared to the signal recorded at 280 K. It cannot be detected in a conventional DNP-enhanced 1H-15N CP spectrum (c) as a broad peak of frozen His-tag at 170 ppm covers both pSB and His75 N2 signals.

ACS Paragon Plus Environment

21

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 29

Fig. 6: (a) DNP-enhanced 13C-CP MAS NMR spectrum of 14,15-13C-all-trans-retinal bound to GPR. A 27-fold signal enhancement is achieved as shown by the spectra recorded with and without DNP. The 13C natural abundance background from protein, lipids and glycerol can be efficiently suppressed by a double-quantum-filter (DQF). (b) Comparison of DNP-enhanced DQF 13C NMR spectra of 14,15-13C-all-trans-retinal in GPR and GPRL105Q. Upon L105Q mutation, the C14 chemical shift changes slightly from 120.2 ppm to 121.3 ppm and the C15 signal shifts significantly from 161.1 ppm to 165.4 ppm.

ACS Paragon Plus Environment

22

Page 23 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Fig. 7: Determination of the retinal conformation at the C14-C15 position by double quantum spectroscopy under DNP-enhanced MAS-NMR conditions. Double-quantum coherence built-up curves for 14,15-13C all-trans-retinal in GPR (a) and GPRL105Q (b) were obtained using POST-C7. The data are well described by 13C-13C dipole couplings of 2665 ± 40 Hz (142 ±1 pm) and 2450 ± 45 Hz (146 ±1 pm) for GPR and GRPL105Q respectively (c). HLF-HCCH dephasing curves for the C14-C15 spin system in GPR (d) and GPRL105Q (e) reporting on the HCCH torsion angle. Subtle differences are observed with the angle changing from 161 ± 3° to 164 ± 2° from GPR to GPRL105Q (f). All the spin dynamics simulations were performed using SIMPSON. The experimentally observed transversal relaxation was accounted for by including a mono-exponentially decaying function in the simulations. The standard deviations of the best-fit solutions are shown in the right panel. See Materials and Methods for further details.

ACS Paragon Plus Environment

23

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 29

Fig. 8: Illustration of the molecular consequences of the L105Q mutation leading to the observed blue shift in GPR. (a) Only subtle structural changes occur as identified by 13C chemical shift perturbations. Their 3D localization is visualized here by different colors using the blue proteorhodopsin X-ray structure (4JQ6), which is shown from the extracellular side 17. Some loops have been omitted. The average chemical shift change is about 0.4 ppm. The most pronounced effect in the retinal binding pocket is observed for the 13C resonances of the T101 side chain, which is in close proximity to L105/Q105 and close to retinal carbons C14 and C15. (b) These local, mutation-induced changes propagate along helix C and into helices D, F and G and affect e.g. T101, W197 and I194 (see Tab. S1). No effects were observed for functionally important residues D97, D227 and K231. (c) The carbonyl group of Q105 points towards C15 and the imine linkage. As a consequence of mutation, the C14-C15 bond length is stretched. QM/MM simulations on retinal show a reconfiguration of the conjugated system between HOMO and LUMO and a shift towards the pSB, which includes an increase of the -orbital character of the C14-C15 bond.66-68 The mutation-induced C14-C15 bondstretching does not alter the bond character in the HOMO but causes an energy increase of the LUMO as the single bond character increases, which results in a shortened conjugated system, a larger HOMO-LUMO gap and therefore in a blue shift. Additional factors include the expected extension of the conjugation effect towards the isomerization region as well as electrostatic effects due to orientation of the polar side chain. HOMO and LUMO of pSB retinal are represented as simplified cartoons as projected from one side of the molecular plane. 67 Opposite signs of the  orbital wave functions are colored in blue and red.

ACS Paragon Plus Environment

24

Page 25 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Table 1: Optical properties of GPR and GPRL105Q in DDM micelles max1

2

(nm)

(nm)

pKa

Primary Reaction3

Photocycle4

(ps)

(ms)

1

2

3

4

t1

t2

t3

t4

t5

KM

M

NO

OPR

OPR

N 5

GPR

520

25

7.05

0.14

0.28

9.5

>1000

0.07

1

5.2

44

210

GPRL105Q

500

37

7.45

0.13

2.3

24

>1000

0.066

0.7

23

83

3800

(1) pH 9; (2) pH4 – pH10; (3) Primary reaction at pH9, global fit analysis of the data shown in Fig. 1d using a sum of exponentials for all wavelengths but different amplitudes; (4) Slower steps of the photocycle at pH9 as determined by flash photolysis in Fig. 1e using a global fit analysis. (5) Data from Hempelmann et al. 13. (6) The time constant does not show a significant amplitude in the spectral range of the K-state (Fig.1e, top), indicating a decay of this photointermediate on a faster time scale.

Table 2: NMR parameters of the chromophore in GPR and GPRL105Q pSB N chemical shift (ppm)

15

GPR GPRL105Q

181 181

13

C Chemical shift (ppm) C14 C15 120.2 121.3

161.1 165.4

C14-C15 dipolar coupling constant (Hz)

C14-C15 bond length (pm)

H-C14-C15-H dihedral angle (degrees)

2665 ± 40 2450 ± 45

142 ±1 146 ±1

161 ±3 164 ±2

ACS Paragon Plus Environment

25

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 29

References (1) Beja, O.; Aravind, L.; Koonin, E. V.; Suzuki, M. T.; Hadd, A.; Nguyen, L. P.; Jovanovich, S. B.; Gates, C. M.; Feldman, R. A.; Spudich, J. L.; Spudich, E. N.; DeLong, E. F. Science 2000, 289, 1902. (2) Hoffmann, J.; Aslimovska, L.; Bamann, C.; Glaubitz, C.; Bamberg, E.; Brutschy, B. Phys. Chem. Chem. Phys. 2010, 12, 3480. (3) Klyszejko, A. L.; Shastri, S.; Mari, S. A.; Grubmuller, H.; Muller, D. J.; Glaubitz, C. J. Mol. Biol. 2008, 376, 35. (4) Beja, O.; Spudich, E. N.; Spudich, J. L.; Leclerc, M.; DeLong, E. F. Nature 2001, 411, 786. (5) de la Torre, J. R.; Christianson, L. M.; Beja, O.; Suzuki, M. T.; Karl, D. M.; Heidelberg, J.; DeLong, E. F. Proc Natl Acad Sci U S A 2003, 100, 12830. (6) DeLong, E. F. Nature 2009, 459, 200. (7) Friedrich, T.; Geibel, S.; Kalmbach, R.; Chizhov, I.; Ataka, K.; Heberle, J.; Engelhard, M.; Bamberg, E. J. Mol. Biol. 2002, 321, 821. (8) Gomez-Consarnau, L.; Akram, N.; Lindell, K.; Pedersen, A.; Neutze, R.; Milton, D. L.; Gonzalez, J. M.; Pinhassi, J. PLoS Biol. 2010, 8. (9) Gomez-Consarnau, L.; Gonzalez, J. M.; Coll-Llado, M.; Gourdon, P.; Pascher, T.; Neutze, R.; Pedros-Alio, C.; Pinhassi, J. Nature 2007, 445, 210. (10) Huber, R.; Köhler, T.; Lenz, M. O.; Bamberg, E.; Kalmbach, R.; Engelhard, M.; Wachtveitl, J. Biochemistry 2005, 44, 1800. (11) Bergo, V.; Amsden, J. J.; Spudich, E. N.; Spudich, J. L.; Rothschild, K. J. Biochemistry 2004, 43, 9075. (12) Reckel, S.; Gottstein, D.; Stehle, J.; Lohr, F.; Verhoefen, M. K.; Takeda, M.; Silvers, R.; Kainosho, M.; Glaubitz, C.; Wachtveitl, J.; Bernhard, F.; Schwalbe, H.; Guntert, P.; Dotsch, V. Angew. Chem. Int. Ed. Engl. 2011, 50, 11942. (13) Hempelmann, F.; Holper, S.; Verhoefen, M. K.; Woerner, A. C.; Kohler, T.; Fiedler, S. A.; Pfleger, N.; Wachtveitl, J.; Glaubitz, C. J. Am. Chem. Soc. 2011, 133, 4645. (14) Shi, L.; Lake, E. M.; Ahmed, M. A.; Brown, L. S.; Ladizhansky, V. Biochim. Biophys. Acta 2009, 1788, 2563. (15) Mehler, M.; Scholz, F.; Ullrich, S. J.; Mao, J.; Braun, M.; Brown, L. J.; Brown, R. C.; Fiedler, S. A.; Becker-Baldus, J.; Wachtveitl, J.; Glaubitz, C. Biophys. J. 2013, 105, 385. (16) Stone, K. M.; Voska, J.; Kinnebrew, M.; Pavlova, A.; Junk, M. J. N.; Han, S. G. Biophys. J. 2013, 104, 472. (17) Ran, T.; Ozorowski, G.; Gao, Y.; Sineshchekov, O. A.; Wang, W.; Spudich, J. L.; Luecke, H. Acta Crystallogr D 2013, 69, 1965. (18) Bamann, C.; Bamberg, E.; Wachtveitl, J.; Glaubitz, C. Biochim. Biophys. Acta 2014, 1837, 614. (19) Man, D.; Wang, W.; Sabehi, G.; Aravind, L.; Post, A. F.; Massana, R.; Spudich, E. N.; Spudich, J. L.; Beja, O. EMBO J. 2003, 22, 1725. (20) Bamann, C.; Nagel, G.; Bamberg, E. Curr. Opin. Neurobiol. 2010, 20, 610. (21) Kelemen, B. R.; Du, M.; Jensen, R. B. Biochim. Biophys. Acta 2003, 1618, 25. (22) Wang, W. W.; Sineshchekov, O. A.; Spudich, E. N.; Spudich, J. L. J. Biol. Chem. 2003, 278, 33985. (23) Amsden, J. J.; Kralj, J. M.; Bergo, V. B.; Spudich, E. N.; Spudich, J. L.; Rothschild, K. J. Biochemistry 2008, 47, 11490. (24) Hillebrecht, J. R.; Galan, J.; Rangarajan, R.; Ramos, L.; McCleary, K.; Ward, D. E.; Stuart, J. A.; Birge, R. R. Biochemistry 2006, 45, 1579. (25) Kralj, J. M.; Bergo, V. B.; Amsden, J. J.; Spudich, E. N.; Spudich, J. L.; Rothschild, K. J. Biochemistry 2008, 47, 3447. (26) Kralj, J. M.; Spudich, E. N.; Spudich, J. L.; Rothschild, K. J. J. Phys. Chem. B 2008, 112, 11770. (27) Maiti, T. K.; Yamada, K.; Inoue, K.; Kandori, H. Biochemistry 2012, 51, 3198. (28) Ozaki, Y.; Kawashima, T.; Abe-Yoshizumi, R.; Kandori, H. Biochemistry 2014.

ACS Paragon Plus Environment

26

Page 27 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

(29) Subramaniam, S.; Greenhalgh, D. A.; Rath, P.; Rothschild, K. J.; Khorana, H. G. Proc Natl Acad Sci U S A 1991, 88, 6873. (30) Haeckel, M.; Hinz, H.-J.; Hedwig, G. R. Biophys. Chem. 1999, 82, 35. (31) Zamyatnin, A. A. Ann Rev Biophys 1984, 13, 145. (32) Wang, S.; Ladizhansky, V. Prog. Nucl. Magn. Reson. Spectrosc. 2014, 82, 1. (33) Carravetta, M.; Eden, M.; Johannessen, O. G.; Luthman, H.; Verdegem, P. J. E.; Lugtenburg, J.; Sebald, A.; Levitt, M. H. J. Am. Chem. Soc. 2001, 123, 10628. (34) Spooner, P. J. R.; Sharples, J. M.; Goodall, S. C.; Seedorf, H.; Verhoeven, M. A.; Lugtenburg, J.; Bovee-Geurts, P. H. M.; DeGrip, W. J.; Watts, A. Biochemistry 2003, 42, 13371. (35) Carravetta, M.; Zhao, X.; Johannessen, O. G.; Lai, W. C.; Verhoeven, M. A.; BoveeGeurts, P. H. M.; Verdegem, P. J. E.; Kiihne, S.; Luthman, H.; de Groot, H. J. M.; deGrip, W. J.; Lugtenburg, J.; Levitt, M. H. J. Am. Chem. Soc. 2004, 126, 3948. (36) Jaroniec, C. P.; Lansing, J. C.; Tounge, B. A.; Belenky, M.; Herzfeld, J.; Griffin, R. G. J. Am. Chem. Soc. 2001, 123, 12929. (37) Etzkorn, M.; Martell, S.; Andronesi, O. C.; Seidel, K.; Engelhard, M.; Baldus, M. Angew. Chem. Int. Ed. 2007, 46, 459. (38) Higman, V. A.; Varga, K.; Aslimovska, L.; Judge, P. J.; Sperling, L. J.; Rienstra, C. M.; Watts, A. Angew. Chem. Int. Ed. 2011, 50, 8432. (39) Wang, S.; Munro, R. A.; Shi, L.; Kawamura, I.; Okitsu, T.; Wada, A.; Kim, S. Y.; Jung, K. H.; Brown, L. S.; Ladizhansky, V. Nat. Methods 2013, 10, 1007. (40) Ni, Q. Z.; Daviso, E.; Can, T. V.; Markhasin, E.; Jawla, S. K.; Swager, T. M.; Temkin, R. J.; Herzfeld, J.; Griffin, R. G. Acc. Chem. Res. 2013, 46, 1933. (41) Bajaj, V. S.; Mak-Jurkauskas, M. L.; Belenky, M.; Herzfeld, J.; Griffin, R. G. Proc Natl Acad Sci U S A 2009, 106, 9244. (42) Jacso, T.; Franks, W. T.; Rose, H.; Fink, U.; Broecker, J.; Keller, S.; Oschkinat, H.; Reif, B. Angew. Chem. Int. Ed. 2012, 51, 432. (43) Ong, Y. S.; Lakatos, A.; Becker-Baldus, J.; Pos, K. M.; Glaubitz, C. J. Am. Chem. Soc. 2013, 135, 15754. (44) Pfleger, N.; Lorch, M.; Woerner, A. C.; Shastri, S.; Glaubitz, C. J. Biomol. NMR 2008, 40, 15. (45) Shi, L.; Ahmed, M. A.; Zhang, W.; Whited, G.; Brown, L. S.; Ladizhansky, V. J. Mol. Biol. 2009, 386, 1078. (46) Yang, J.; Aslimovska, L.; Glaubitz, C. J. Am. Chem. Soc. 2011, 133, 4874. (47) Sharaabi, Y.; Brumfeld, V.; Sheves, M. Biochemistry 2010, 49, 4457. (48) Kiihne, S. R.; Creemers, A. F. L.; de Grip, W. J.; Bovee-Geurts, P. H. M.; Lugtenburg, J.; de Groot, H. J. M. J. Am. Chem. Soc. 2005, 127, 5734. (49) Yao, X. L.; Schmidt-Rohr, K.; Hong, M. J. Magn. Reson. 2001, 149, 139. (50) Song, C. S.; Hu, K. N.; Joo, C. G.; Swager, T. M.; Griffin, R. G. J. Am. Chem. Soc. 2006, 128, 11385. (51) Verhoefen, M. K.; Schafer, G.; Shastri, S.; Weber, I.; Glaubitz, C.; Mantele, W.; Wachtveitl, J. Biochim. Biophys. Acta 2011, 1807, 1583. (52) Hamanaka, T.; Kakudo, M.; Ashida, T.; Mitsui, T. Acta Crystallogr Sect B 1972, B 28, 214. (53) Concistre, M.; Johannessen, O. G.; McLean, N.; Bovee-Geurts, P. H.; Brown, R. C.; Degrip, W. J.; Levitt, M. H. J. Biomol. NMR 2012, 53, 247. (54) Lansing, J. C.; Hohwy, M.; Jaroniec, C. P.; Creemers, A. F. L.; Lugtenburg, J.; Herzfeld, J.; Griffin, R. G. Biochemistry 2002, 41, 431. (55) Yoshizawa, S.; Kumagai, Y.; Kim, H.; Ogura, Y.; Hayashi, T.; Iwasaki, W.; DeLong, E. F.; Kogure, K. Proc Natl Acad Sci U S A 2014, 111, 6732. (56) Kloppmann, E.; Becker, T.; Ullmann, G. M. Proteins: Struct. Funct. Bioinform. 2005, 61, 953. (57) Mathies, R.; Stryer, L. Proc Natl Acad Sci U S A 1976, 73, 2169. (58) Schenkl, S.; van Mourik, F.; van der Zwan, G.; Haacke, S.; Chergui, M. Science 2005, 309, 917. (59) Yan, B.; Spudich, J. L.; Mazur, P.; Vunnam, S.; Derguini, F.; Nakanishi, K. J. Biol. Chem. 1995, 270, 29668. ACS Paragon Plus Environment

27

Journal of the American Chemical Society

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 29

(60) Hu, J. G. G.; Sun, B. Q. Q.; Petkova, A. T.; Griffin, R. G.; Herzfeld, J. Biochemistry 1997, 36, 9316. (61) Hu, J. G. G.; Griffin, R. G.; Herzfeld, J. J. Am. Chem. Soc. 1997, 119, 9495. (62) Buda, F.; deGroot, H. J. M.; Bifone, A. Phys. Rev. Lett. 1996, 77, 4474. (63) Verhoeven, M. A.; Creemers, A. F. L.; Bovee-Geurts, P. H. M.; De Grip, W. J.; Lugtenburg, J.; de Groot, H. J. M. Biochemistry 2001, 40, 3282. (64) Creemers, A. F. L.; Bovee-Geurts, P. H. M.; DeGrip, W. J.; Lugtenburg, J.; de Groot, H. J. M. Biochemistry 2004, 43, 16011. (65) Melaccio, F.; Ferre, N.; Olivucci, M. PCCP 2012, 14, 12485. (66) Fujimoto, K.; Hayashi, S.; Hasegawa, J.-y.; Nakatsuji, H. J Chem Theor Comp 2007, 3, 605. (67) Fujimoto, K. J.; Asai, K.; Hasegawa, J.-y. PCCP 2010, 12, 13107. (68) Lee, H. M.; Kim, J.; Kim, C. J.; Kim, K. S. J. Chem. Phys. 2002, 116, 6549. (69) Subramaniam, S.; Henderson, R. Nature 2000, 406, 653. (70) Lanyi, J. K.; Luecke, H. Curr. Opin. Struct. Biol. 2001, 11, 415. (71) Hirai, T.; Subramaniam, S. PLoS One 2009, 4, e5769. (72) Vonck, J. EMBO J. 2000, 19, 2152. (73) Kuhlbrandt, W. Nature 2000, 406, 569. (74) Kim, S. Y.; Waschuk, S. A.; Brown, L. S.; Jung, K. H. Biochim. Biophys. Acta 2008, 1777, 504. (75) Ward, M. E.; Shi, L.; Lake, E.; Krishnamurthy, S.; Hutchins, H.; Brown, L. S.; Ladizhansky, V. J. Am. Chem. Soc. 2011, 133, 17434. (76) Yao, X. L.; Hong, M. J. Biomol. NMR 2001, 20, 263. (77) Vinogradov, E.; Madhu, P. K.; Vega, S. Chem. Phys. Lett. 1999, 314, 443. (78) Fung, B. M.; Khitrin, A. K.; Ermolaev, K. J. Magn. Reson. 2000, 142, 97. (79) Hohwy, M.; Jakobsen, H. J.; Eden, M.; Levitt, M. H.; Nielsen, N. C. J. Chem. Phys. 1998, 108, 2686. (80) Bak, M.; Rasmussen, J. T.; Nielsen, N. C. J. Magn. Reson. 2000, 147, 296. (81) Hu, J.; Griffin, R. G.; Herzfeld, J. Proc Natl Acad Sci U S A 1994, 91, 8880.

ACS Paragon Plus Environment

28

Page 29 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

TOC Graphic

ACS Paragon Plus Environment

29